Who needs redundancy?

FLASH! Telecom and the web are merging together. OK – so I’m the master of the obvious. But as these two world collide, I see a sharp contrast in thinking.

Yesterday, Blackberry service was 100% disrupted. No word on why. My guess is a telecom guy was involved here somewhere. In meeting with IP guys and a traditional telecom guys how they build and construct things is totally different. I like the IP guys attitude.

Telecom guys spent a lot of time telling me about how redundant their systems are. Protected from mal-formed messages and bad stuff. It sounds like they build a castle, dig a moat around it, fill it with water, populate it with some gators and call it a day. They can talk for hours about what happens when "something goes wrong" and how they continue to work. They’re smart you know, but everytime something goes wrong, it’s always someplace they didn’t think about

I contrast this with IP guys, who seem oblivous to all the hard work that’s been done in the telecom industry. Their attitude is simply, yup, this baby likely to blow up because we wrote some crappy code somewhere, but so what, some other server somewhere in the network will just keep chugging along, yeah you might lose something, but it’s so distributed it will only hurt and few and who cares.

I like that attitude. I’ve been reading the many public documents on how the Google search engine was built. Cheap off the shelf hardware, an understanding that stuff would simply fail, where, when & why – wasn’t that important, a focus on ensuring that any single outage out not roll through the other systems and impact them.

I’m afraid the days of talking about A/B side, active/standby, N+1 redundancy may well be passing us by.